Television – Object tracking
Reexamination Certificate
1998-05-14
2002-06-11
Kelley, Chris (Department: 2613)
Television
Object tracking
C382S219000, C382S218000, C382S103000
Reexamination Certificate
active
06404455
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates to a monitoring apparatus provided with a TV camera and more particularly to a method for tracking an object and an apparatus for tracking and monitoring an object which are both arranged to automatically detect one or more objects entering or intruding in a field of view of a TV camera from a video signal sent from the TV camera and to track movement of the detected object(s) or magnify the object(s) on a TV monitor.
A manually operated picture monitoring apparatus arranged to use a TV camera has been prevailing in the past. In recent years, however, in place of monitoring the object on the monitor by a person, the type of monitoring system has been demanded which automatically detects a moving object such as a person or a vehicle entering a monitoring field of view from a video signal and display the object in magnified representation or give a certain notice or an alarm.
To detect a moving object or an entering object, a conventional system has been arranged to compare an input video signal obtained from the TV camera and a reference background video signal, that is, a video signal that does not contain an object to be detected, derive a difference of a brightness value at each pixel between both of the video signals, and detect an area with a large difference as an object. This system is called the difference method, which has been widely used. The application of this system is described in U.S. patent application Ser. No. 08/646,018 filed on May 7, 1996 (corresponding to the European Application No.963033030 filed on May 13, 1996). Another application of the tracking system with the difference method is discussed in the thesis entitled “A procedure for the Analysis of Skiers' Movement with Image Processing Techniques”, Vol. 43, No. 12, pp. 1370 to 1374 of “Proceedings of TV Society” issued on December, 1989.
One of plausible methods arranged to automatically detect an entering object from a video signal obtained from a TV camera, track the entering object and magnify it on a video monitor screen for the purpose of facilitating the monitor of the entering object is to detect the entering object by successively switching the video signals outputted from a plurality of TV cameras. However, this method disadvantageously involves many TV cameras and a video signal switcher and a controller therefor so that this method makes the overall system larger in scale.
For decreasing the necessary number of TV cameras for the purpose of reducing the system scale, the system is required to widen a field of view of an imaging lens provided in each TV camera. This system however makes the entering object appear on the monitor in a small size. For example, if the entering object is a person, this system may have difficulty in discriminating the lineaments of the entering person.
Further, when an object having a luminance value close to the background picture is to be detected by the difference method, a part of the object may be lost undetected or the object may be split into a plurality of pieces, resulting in detecting the object as a plurality of objects. If one object within a field of view is detected as a plurality of objects, therefore, the method for tracking an object using the difference method may determine that the same number of objects as the detected member of the objects exist and track the plurality of the objects though in actuality just one object exists within the field of view. Further, the tracking method using the difference method detects objects that are seemingly connected together and looks as one object as one object. In a case that, therefore, a plurality of objects exist within the field of view, if these objects are connected together at a certain point in time, this type of method determines that there exists only one object at the time and hence track them as one object though in actuality two or more objects exist within the field of view. Moreover, since the tracking method using the difference method involves no comparison of shapes of a picture nor no comparison of picture patterns, if two or more objects are located within the field of view, this method can not reliably guarantee that the same object is being tracked.
SUMMARY OF THE INVENTION
It is an object of the present invention to provide a picture processing system for tracking an entering object by automatically detecting the entering object and magnifying the detected entering object, for allowing an observer to more easily make sure of the entering object on a video monitor screen.
In carrying out the object, a picture processing system for tracking an entering object according to an aspect of the invention includes a camera or a TV camera with an imaging lens of a wide field for outputting a video signal used for detecting the entering object, another camera or TV camera with a zoom lens for outputting a video signal used for monitoring the picture as tracking the entering object, the camera being mounted on an electric pan and tilt head or pan head, and a picture processing device for processing and operating a video signal inputted from the camera provided with the lens of the wide field and outputting a rotation control signal for rotating the electric pan and tilt head by a necessary amount of angle, wherein the object entering the predetermined monitoring area is automatically tracked and picked up. Here, the camera or TV camera provided with the imaging lens of a wide field is a camera or a TV camera which is capable of imaging the whole of the picture within the object tracking and monitoring area. The whole of the object tracking/monitoring area is referred to as the wide field or field of view.
A system for processing a tracking picture of an entering object according to another aspect of the invention includes a camera or a TV camera with a lens of a wide field for outputting a video signal used for detecting the entering object, another camera or TV camera with a zoom lens for outputting a video signal used for monitoring a magnified picture as tracking the entering object, the camera being mounted on an electric pan and tilt head, a picture processing device for processing and operating the video signal sent from the camera with the lens of the wide field and outputting a rotation control signal for rotating the electric pan and tilt head by a necessary amount of angle and a zooming control signal for zooming the zoom lens by a necessary amount, wherein the object entering the predetermined monitoring area is automatically tracked and picked up.
Preferably, the picture processing device in the system includes a picture memory for storing a video signal sent from the camera with the lens of the wide field and a processed signal of the video signal, a CPU (central processing unit) for carrying out processing, operation, control and management performed in the picture processing device, a program memory for storing a program for operating the CPU, a working memory for storing data for analyzing data obtained by processing a locational coordinate and a movement history of the detected entering object, and a pan and tilt head control unit for outputting the rotation control signal for rotating the electric pan and tilt head by a necessary amount of angle and the zooming control signal for zooming the zoom lens by a necessary amount based on the operated data.
Preferably, the picture memory provided in the picture processing device includes an input picture memory for storing an inputted video signal, a background picture memory for storing a background video signal used for detecting the entering object, two working memories for storing the background video signal and the input video signal for use in carrying out an operation of detecting a difference between the pictures, a picture memory for storing an inter-picture operation result obtained by detecting and operating a difference between the background picture and the input picture, and a binary-coded picture memory for storing a binary-coded video signal obtained by op
Endo Miyuki
Ito Wataru
Okada Toshimichi
Ueda Hirotada
An Shawn S.
Antonelli Terry Stout & Kraus LLP
Hitachi Denshi Kabushiki Kaisha
Kelley Chris
LandOfFree
Method for tracking entering object and apparatus for... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for tracking entering object and apparatus for..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for tracking entering object and apparatus for... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2951093